I went outside this morning with my large protractor and marked the angle
of the building verses my plotted north line. Everything agrees within
one
degree, which is better than I can eyeball Polaris against a point on the
building to a rock on the ground. I am "square" with the street and one
neighbor's fence. The other neighbor's fence is at an obvious angle.
I ran outside at 12:55pm PDT to find the solar noon marking of north.
I had put a couple of rocks down last night, and noted the one tree.
Today, with the sun shining, I tried to put more rocks down... Polaris
doesn't move much, but solar midday is a fleeting opportunity.
By the time it occurred to me to let a plumb bob string cast a shadow on
the
line I had drawn on the deck from the Polaris lineup, it was obviously
off kilter, and moving away quickly.
I also tried lining up the tree and the point on my deck from farther out,
and projecting a point with my GPS due south. That isn't particularly
satisfying. The GPS started out at an estimated accuracy of 117 feet, and
settled in at 17 feet.
The GPS is wonderful for getting within sight of an object, but it isn't
great for survey work.
http://www.confluence.org/photo.php?visitid=8914
picture 5 shows someone else visiting my pile of rocks in the middle of a
large open plateau.
I'm going to chisel a north marker in the sidewalk ;-)